Understanding Dynamic Balancers: redux

[JMarkWolf]

Thanks to Greg Locock and Klyde for their posts in the original thread.

I have since made progress in my quest to understand the workings of the typical propeller/rotor dynamic balancer, but the original thread was closed, so I started a new thread.

I now have a desktop "rotor simulator" fabricated with a small variable speed AC motor, and an 8" flywheel to which I can attach trial weights at 10 degree increments. My simulator sounds much like Klydes' description of his.

Capturing scope-screens and Excel files from the accelerometer and Hall effect phase marker is now a snap.

The accelerometer signal is a beautiful, nearly noiseless sine wave. The phase markers are crisp and clean.

Initial data that I've captured, and rudimentary waveform analysis I've done, seems to indicate that the positive going velocity zero-crossing is different than the phase angle reported by the commercial instrument to which my simulator is attached.

The poccess I'm using is to generate an integral (velocity) curve of the acceleration waveform, via Sigview waveform analysis tools. The positive going zero-crossing of the velocity curve corresponds to 162 degrees from the zero degree phase marker.

The commercial instrument reports that phase angle at 320 degrees. A difference of a factor of 2. I'll test other samples at other imbalance to see if this holds true.

I've measured the phase lag "stack-up", due to electronics circuitry, of the commercial instrument between the input connector and the output of the anti-aliasing filter, and it appears to be precisely (and happily) 360 degrees. So electronics-induced phase lag error appears that it should not be a factor here.

The 90 degree shift due to critical speed that Klyde described in his post doesn't seem to apply here, although I do not yet know what the critical speed of my rotor simulator is.

I'd be happy to supply screen captures or Excel files if anyone cares to see them.

Can anyone offer some insight here?

 

[GregLocock]

Where is the trial mass? At zero degrees?

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[JMarkWolf]

Yes, the trial weight (a short stack of washers under the head of a 6-32 screw) is currently at the zero degree mark of the platter.

The signal from the Hall effect switch transitions right at this zero degree mark.

The actual weight is unknown. My digital postage meter is not sensitive enough to measure it.

 

[GregLocock]

First thing to do is check that the system is properly balanced, or at least that your trial mass produces a signal 10 times larger than the baseline.

I'm not neccesarily convinced that velocity is the obvious signal to use, I've always used acceleration.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[JMarkWolf]

Without the trial weight, the commercial instrument reports 0.00ips (phase is irrelevant at this point).

On further waveform analysis, I've learned that the neither the peaks nor the zero crossings of either the accel or velocity waveform correspond to the phase angle indicated by the commercial instrument.

 

[GregLocock]

Hang on is the commercial unit a balancing system, or a phase meter?

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[JMarkWolf]

It's a balancing instrument, hence the title of this thread, however it provides phase angle as part of it's "report" on it's 1-line alphanumeric LCD ("0.75ips @ 307 deg" typical).

Phase angle "from what" is currently the big question.

 

[GregLocock]

OK, does it come with a balancing program?

If so just try a balance with it, you'll soon see if it is making sense.

It does sound odd, I'd agree.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[JMarkWolf]

No, it has no accompanying "program". It is a completely self-contained hand-held instrument, except for the polar charts, on which you manually "plot the dot", specified by the instrument on the integral LCD.

The location of the dot on the polar charts, points to the specific corrective action required to reduce the imbalance and move the "next dot", acquired during the next acquisition, closer to the center of the chart.

I'll post an inquiry on the Vibration forum.

 

[thruthefence]

With some balancers I'm familiar with, working on aircraft, not experimentally, you would make a weight change, & observe the "move line" the weight change caused. If it didn't head right thru the center of the "clock", then you had a process called "correcting the clock angle" where you added or subtracted 'hours' from the clock chart, and continued with the balance procedure.

If I could segue off track a little, Greg, I noted your comment on IPS vs velocity, and I have also noticed the vibration pick-ups that were once described as "accelerometers" are now called "velometers" , at least with one system I'm familiar with. Is there a difference?
regards, Steve

 

[JMarkWolf]

thruthefence wrote:

"With some balancers I'm familiar with, working on aircraft, not experimentally, you would make a weight change, & observe the "move line" the weight change caused. If it didn't head right thru the center of the "clock", then you had a process called "correcting the clock angle" where you added or subtracted 'hours' from the clock chart, and continued with the balance procedure"

This procedure sounds like the process for creating the polar charts for a new machine.

I already have the polar charts for my aircraft.

 

[GregLocock]

I'd guess that zero degrees for the commercial instrument would be the rotational speed marker.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.